fraiser

fraiser

This is largely stolen from Jenny Bryan and the handout for stat 545. If you want the much better handout just google happygitwithR

What is Git and why should I learn it?

Git is simply a version control system, with the original purpose to help developers work together on software projects. It has it’s own coding structure but for the purpose of this handout, we are going to be using the easy version.

This is done by using repositories or repos which are basically highly structured super powered team drive. The data science world has taken it a new level because of that open-science thingy.

Simply put, it allows you to functionally keep control of multiple edits while working with other people

What are the benefits to it?

Exposure If someone wants your code, they can just clone it from your Github.

Help each other Let’s say I want to advance someone else’s code. I can fork a feature or fix bugs for them. Github will automatically notify the author and they can use keep or reject edits.

Why can’t I copy and paste in a word file? Or just send people my R code?

Because it’s sloppy, nobody will hire you, and nobody likes having fifteen files called Version 1, version 2 final Version, The actual final version.

People are afraid of change, think of why do we use gmail instead of writing letters and putting money on the letter and sending each other. Technology advances and people who take advantage of it waste less time.

What is Github?

GitHub is a website that hosts your data, but it also has a software component that is very powerful and can be accessed directly through your code in R. So in a way yes, Github is like DropBox or Google Drive.

There are plenty of alternatives to GitHub, but we are choosing to use this one because they have seemingly dominated the Git Field.

How do I get GitHub? Is it easy?

Google Github.com and go to their website and make an account, while you are there download the software client for it.

Good Job. Though, I want to mention here that there are MUCH better clients for Git AND many ways to utilize Git.

Step 1. Create a repository.

A repository is essentially a folder for a single project. I am going to be creating a Github project called IntroToGit which is a Rmarkdown file in which I explain how to use Github for a Regression class at Humboldt State University.

We start this by selecting new repository and just selecting a path or folder for you.

Nice! Fill your repository with nice and shiny things. At some point, you’ll want to publish the repository so it can go up on your account on the github website.

For example, my repository consists of the rmarkdown file, the html file for it, and some screenshots for you to follow along. Actually, exactly this picture below.

Before you can publish you need to do three things.

Once you have published, the whole world can see your repository!

Branches

Alright, so Let’s talk about branches, These are the aspects for collaboration. Instead of always working in one specific master file and messing that up, we can work in branches and discuss whether or not we want changes.

We use branches for features or edits or updates that we can experiment with before adding them into the master file. This is useful because you won’t have those final version seriously final version problems.

How do I do this? Just go to the branche section in the GitHub client and create a new branch. I will be naming mine Gif because I want to experiment with this markdown file by adding gifs to it.

Make sure you are in the proper branch before experimenting.

So here we are in the Gif Branch. I want to add a gif to the title because I think it’s a great idea.

This is way too meta
Don’t ask me how I did this, you’ll…make fun of me

Now add a definition to your Github and Commit!

Now you have two branches, you can merge both branches into the master. That is, if you are happy with it.